﻿238 Profs. Liveing and Dewar on the Spectrum of 



seen in the glow about both poles, and equally well when the 

 electrodes were cold, and when they were hot, even red-hot. 

 There was a third much fainter band in the orange a little 

 less refrangible than the D lines of sodium. There is no 

 doubt that these bands are the same as have been described 

 by Schuster in the negative glow in a vacuous oxygen-tube. 



When a jar was put in the circuit, these bands disappeared 

 or nearly so. While one electrode remained in the liquid, a 

 good many bright lines came out on putting the jar into the 

 circuit. One of these was a line at wave-length about 557, 

 the same as previously seen, nearly in the middle of one of 

 the green bands. In fact, it seemed to replace the band when 

 the jar was put into the circuit. It was not, however, so 

 permanently seen as some of the other lines. Only a few of 

 these other lines were measured. One appeared to be an 

 oxygen line at about X 544 ; a pair at about X 566 may have 

 been air-lines as the oxygen usually contains a little air. 

 Another was a platinum line about X 455. 



The line above mentioned, X 557, appears to be of some 

 interest because it falls very near the place of the auroral line, 

 and the conditions under which it is produced resemble, in 

 regard to the low temperature, and to some extent in regard 

 to the pressure, those in which Auroras are produced. 

 The place of this line was measured several times, but the 

 circumstances of these experiments were not very favourable 

 for exact measurements. The measure which we marked as 

 the best gave a wave-length for the line 5572, but other 

 measures gave figures between that and 5578. It seems, 

 therefore, not improbable that this line may be identical with 

 the auroral line, of which the exact wave-length cannot even 

 now be said to be quite certainly determined, but which has 

 probably the value 5571. 



The identity of the line we have observed with the auroral 

 line cannot be said to be proved as yet. Further observation 

 is needed, and we hope to carry our experiments further. 

 The broad green bands do not seem to be connected with a 

 low temperature, but were produced by the discharge without 

 a jar ; this line was, so far as we observed, only produced 

 when one electrode was immersed in the liquid, and therefore 

 cold, and when a jar was in circuit. 



The passage of the discharge through the liquid produced 

 much ozone. Not only was the smell of ozone very strong, 

 but the liquid took the indigo tint, deeper than the blue of 

 ordinary oxygen, which is characteristic of ozone. On one 

 occasion, after the sparks had been passed through the liquid 

 for a short time, an explosion ensued which shattered the 



